Dividend Policy and Investment Efficiency: Analysis of the Moderating Effect of the Quality of Financial and Accounting Information of Companies Listed on the West African Stock Market: Case of BRVM ()
1. Introduction
Investment, financing and dividend distribution policies define the financial policies of company directors (Morissette, 2008). Of these, the investment decision is seen as the most important for any company director, since investments contribute to the growth and wealth creation of companies and constitute their raison d’être. This is why the company is seen as a set of investment projects.
Charreaux (1991), defines the investment decision as any decision that leads to the purchase of an asset with the aim of obtaining future cash flows in order to maximise the wealth of the owners. To expand on this definition, (Morissette, 2008) considers that the investment decision is any decision that leads to the acquisition of tangible or intangible assets that can ensure the company’s growth. In the same vein, Laopodis, 2020 states that investment is an allocation of current resources (intelligence, labour, capital, etc.) that consists of purchasing or exchanging other resources with the aim of generating profits.
Chaleeda et al. (2019) believe that managing these profits leads companies to adopt a dividend distribution policy, such as proportional distribution, full retained earnings for shareholders or reserving all profits for new investment opportunities. The more a company retains these profits for reinvestment, the fewer dividends are paid and vice versa. Dividend policy is therefore one of the most controversial policies in corporate finance.
According to some authors, dividend policy is seen as a signal that companies send to the market. Dividend payments signal the company’s prospects, because they attract the interest of shareholders and potential investors and thus affect the company’s future growth.
Dividends are a way for some companies with a strong reputation to reassure shareholders by offering a large amount from net profits. Others, on the other hand, prefer to finance themselves from cash flow, which deprives shareholders of the dividends they receive. This situation is often a source of conflict between the principal and the agent, and is part of two opposing theories.
The dividend irrelevance theory considers that there is no significant relationship between dividend payments and the company’s potential profitability or share price. The dividend relevance theory, on the other hand, proves a correlation between dividend policy and companies’ investment decisions.
Pioneering studies carried out in the European context have produced mixed results. Miller and Modigliani (1961) and Smirlock and Marshall (1983) proved that in the presence of a pure and perfect market, with transparency of information, the relationship between investment decisions and dividend policy is irrelevant. On the other hand, studies by Harakeh (2020) and Nguyen (2022) found a negative and relevant relationship between dividend policy and investment decisions.
The West African stock market, through its regulatory body, the Bourse Régionale des Valeurs Mobilières (BRVM), has seen unprecedented growth in recent years. With an equity market capitalisation of FCFA 8000 billion, and respective growth in the BRVM 30, Prestige and composite indices of 7.85%, 1.47% and 5.38%, this financial institution ranks among the top five stock exchanges on the African continent, and eighth in terms of market capitalisation at continental level. In 2023, there will be nearly fifty-eight companies listed on the BRVM, an increase on previous years. Companies listed on the stock exchange make an enormous contribution to the continent’s economic growth through the drainage of capital and investment. They account for 60% of the continent’s GDP and employ more than 500 people. Companies listed on the BRVM are also of great importance from a tax point of view, with more than 1.3 billion CFA francs in tax paid. Unfortunately, these companies face enormous difficulties in terms of their financial policy, particularly their investment, financing and dividend policy.
Given this observation, the lack of studies on this topic at the BRVM and the various results of research on European stock markets, it is urgent to examine the influence of the quality of accounting information in the relationship between dividend policy and investment efficiency on the West African stock market. This research aims to assess the extent to which dividends influence the investment efficiency of companies listed on the BRVM from 2011 to 2023 by analysing the quality of accounting information as a moderating variable.
Our study is divided into three sections. The first deals with the literature review on the subject, the second with the methodology adopted, and the third with the presentation and analysis of the results.
2. Literature Review and Hypothesis Development
Agency, asymmetric information and signal theories form the theoretical underpinnings of our study, which analyses the quality of accounting information as a moderating variable in the relationship between dividend policy and the investment decision. These different theories are the subject of our study.
2.1. The Agency Theory of Jensen and Meckling (1976)
The existence of agency theory stems from the multiplicity of categories of participants in an organisation and the resulting divergences in contracts. These divergences generate conflicts between the principal and the agent. Managers must then reconcile the different interests of shareholders to resolve conflicts of interest between agents and principals. As Harakeh (2020) states, the presence of managers in the entity has accentuated the divergences between agent and principal. The latter demand a high return from a high dividend rate, whereas managers prefer a low dividend rate in order to take advantage of the profits set aside to finance their investment projects. In this situation, managers are more flexible in making investment decisions and use more internal resources than external ones, which have a relatively high cost.
Because of their position and the bonuses they receive, executives can over-invest in projects without the consent of shareholders, which is often a source of information asymmetry problems. In the same vein, Jensen, M. C., & Meckling, W. H. (1976) assert that managers tend to adopt a less optimal investment policy since they benefit from the company’s reputation based on its size, power and high returns. To remedy this situation, dividend payments are used as a solution to the problems of divergence between the principal and the agent. This means that dividend payments can reduce the excess resources available within the company to avoid overinvestment, which arises from the problem of asymmetric information.
2.1.1. The Theory of Asymmetric Information
Market inefficiencies caused by adverse selection and moral hazard lead company managers to deviate from their optimal investment level, either by over- or under-investing. In this context of asymmetric information, company managers always have the possibility and an overflowing imagination to show in the accounts a very advantageous financial situation insofar as the investment decision is their responsibility, they most often proceed to a dosage of the invested level according to their own objectives.
According to Healy and Whalen (1999), managers use their judgement in financial reporting and in the verbiage of the accounts to truncate financial reports with the aim of either misleading the various stakeholders about the economic results, or modifying the contractual results which are based on the declared accounting figures.
The African stock market, and in particular the Bourse Régionale des Valeurs Mobilières (BRVM), is an emerging market that faces many difficulties, including asymmetric information. This situation is at the root of the relatively high cost of external capital for listed companies. As stated by Myers & Majluf (1984), asymmetric information increases the cost of external capital, forcing managers to prefer their own resources to external financing. The payment of dividends thus reduces the liquidity available within companies and influences economic decisions regarding investment.
2.1.2. Signal Theory
The latter sees the dividend as a way for the company to communicate with the market. By increasing dividends, the company sends out a signal about its future earnings performance and reassures investors. As M’rabet, Rachid, and Wiame Boujjat (2016) point out, a company that maintains its dividend level despite a fall in profits is signalling to the market that this fall is structural and that profits will continue to rise. Thus, Dang et al. (2021) confirm that dividend signalling indicates that the company can obtain good investment opportunities and transmits good information to its investors to make profits.
The various models of the signal constituted by the payment of dividends, such as Bhattacharya (1979), John and Williams (1985) and Miller and Rock (1985), show that managers have more privileged information about the value of the company than investors, and they use dividends to transmit this information to the market.
In an imperfect market, increasing dividend payments means that shareholders anticipate that the company’s future earnings should increase. Similarly, a reduction in dividend payments may alert shareholders to the possibility of a fall in future profits, which could alter the value of the company. This justifies the conclusion of Brav et al. (2005) that it is difficult for companies to reduce dividend payments and planned investments. Signal theory can therefore provide explanations for the impact of dividend policies on investment decisions
2.2. Previous Contributions
2.2.1. Investment Decisions and Accounting Information
The purpose of all data is to present a set of details, numerical facts and explanations to facilitate accurate analysis of the situation and more confident decision-making. For Sauviat (2002), information is the key element in the global economy. All decision-making in an organization is based on the information available about it. The accounting information system plays a vital role, even if accounting was previously considered an esoteric technique, the preserve of professionals and a tax issue. Accounting is now a strategic steering and decision-making tool (Brigitte et al., 2010). Its purpose is also to provide information on the company’s relations with the outside world.
Verdi (2006), studying the link between the quality of accounting information and investment efficiency, finds that high quality accounting information is negatively associated with both overinvestment and underinvestment. He also states that the quality of accounting information reduces underinvestment more strongly when firms face financing constraints. Similarly, it reduces overinvestment much more for firms with a large cash position.
The reliability of accounting information contributes enormously to the reduction of financial resources subject to overinvestment and positively increases financial expenditure relating to underinvestment. This justifies the work of Biddle et al. (2009) showing the considerable influence of the quality of accounting information on the effectiveness of investment policies through overinvestment and underinvestment by companies.
Recent studies on the Nigerian market have also investigated the link between the quality of accounting information and investment policies. As shown in the studies by Nwaobia et al., (2016), Mohamed & Schwienbacher, (2016), Zayol, Agaregh and Eneji (2017); Sylvia Fettry and Tita Djuitaningsih (2019) and recently which unanimously conclude, the positive impact of the quality of accounting and financial information on investment decisions and the positive reaction of the stock market.
By studying the link between the effectiveness of the investment policy and the quality of accounting information on the African stock market, our hypothesis can be formulated as follows
H1: The quality of accounting information influences investment policy on the BRVM
2.2.2. Relationship between Investment Policy and Dividend Policy of
BRVM Companies
Decisions on whether or not to pay dividends within a company are very important. Dividends are a way for shareholders to control investment decisions and reduce opportunistic behaviour by management.
The distribution of dividends is often adopted by certain companies to reduce the discretionary power of management and avoid the allocation of cash resources to high-risk projects. The adoption of a rate of profit redistribution by companies reduces the problems associated with cash flow management and contributes to the reliability of accounting and financial information. (Houewou, 2022).
Many studies of the link between investment policy and dividend policy have produced divergent results.
Saddiq et al. (2018), testing the effect of dividend distribution on investment in Nigeria, finds a positive effect between dividend policy and investment-related policy.
For Uzomah and Ihe (2021), companies need to maintain a stable dividend payout pattern. Thus, if managers pay too many dividends, companies may lack the liquidity to seize growth opportunities. Conversely, underpaying dividends can lead to shareholder dissatisfaction and free cash flow problems (Jensen & Meckling, 1976). However, managers must then compromise on dividends in order to pursue their business objectives.
The decision to distribute dividends also depends on the company’s position in its sector of activity. Fast-growing companies need financial resources to be able to position themselves better and face up to the competition. Faced with this strategy, they opt not to distribute dividends (La Porta et al., 2000).
Nkuah & Yusif (2016), conducting a study on the Ghanaian stock market, revealed that dividend payments have a positive and significant impact on the market price stock. They stipulate that the impact is more accentuated at the level of companies that distribute dividends than those that retain them.
On the other hand, Nguyen (2022), conducting a study on the stock market in Vietnam, concluded that dividends have a negative and significant effect on investment decisions, since companies retain more liquidity to make their investments and pay fewer dividends. Brav et al., (2005) and Ramalingegowda et al., (2013) also found a negative relationship between dividend and investment policies. They argue that dividend payouts can lead to underinvestment. In the same vein, Kehinde, Uwalomwa, Olubukola, Osarienne and Sylvestre (2017) in their work on Nigerian banks, came to a negative result, thus stipulating that dividend yield and retention rates had a negative and significant effect on the market price per share. Thus, they recommend that Nigeria should favour a high dividend payout ratio to value its shares on the Nigerian stock market.
Sahibzada and Zubair (2017) in their study of technology companies in the US, found that dividend policy does not influence investment decisions.
By analysing the influence of dividend policy on the investment of companies listed on the African stock exchange, our hypothesis can be formulated as follows:
H2: Dividend payouts influence the investment policy of BRVM-listed companies
2.2.3. Quality of Accounting Information as a Moderating Variable
Financial and accounting information is important for decision-making within a company. Of these decisions, the one relating to the distribution of profits is crucial. To maintain his position within the company, the director must be able to arbitrate between the satisfaction of the owners and the value of the company in terms of the profitability of the shares and the effectiveness of the investments to be made.
According to Bushman et al., (2001), quality information is valuable for cash flows and investment projects. It reduces the problem of asymmetric information between management and investors, as investors are less likely to buy shares at a high price.
High-quality accounting and financial information improves investor control by reducing the difficulties associated with moral hazard in the signing of contracts between investors and management. Harakeh (2020) and Ramalingegowda et al. (2013) confirm that companies with better quality accounting information have lower debt and capital costs than other companies. This means that companies with high-quality accounting information can better access external funds or raise capital at a lower cost.
High-quality accounting information enables companies to reassure investors while maintaining their dividend policy. Investors are committed to supporting the company’s dividend policy when there is transparency and reliability of financial data and also improved financial profitability. This situation allows investors to better manage their capital by effectively improving investment decisions.
Quality accounting information then helps investors to better manage their capital by effectively improving investment decisions. In view of the above, we can stipulate that the quality of accounting information can be understood as a moderating variable in the link between dividend policy and the effectiveness of investment decisions. Our hypothesis can be formulated as follows:
H3: the quality of accounting information moderates the relationship between dividend policy and investment efficiency in BRVM-listed companies
Based on the above assumptions, the conceptual model of our study is as follows in Figure 1:
Figure 1. The conceptual research model. Source: Authors.
3. Research Methodology
3.1. Sources of Data Collected and Processing Methods
The aim of this quantitative study is to investigate the effect of the quality of accounting information on the relationship between dividend policy and investment efficiency. To do so, we collected information on companies listed on the Bourse Régionale des Valeurs Mobilières from 2011 to 2023. We believe that this data source is credible because, under the rules governing the WAEMU stock exchange, all listed companies are required to submit audited financial statements.
Up to 2023, there will be 58 companies listed on the BRVM, including twelve (12) in the finance and insurance sector, five (05) in the utilities sector, twenty-one (21) in industry, six (06) in agriculture, eight (08) in distribution, two (02) in transport, three (03) government bonds and one (01) other sector. For our study, we have excluded companies in the finance and insurance sector as well as those relating to government bonds. Companies for which not all reports are available are also excluded.
Our sample is as follows in Table 1:
Table 1. Selection of the study sample.
|
BRVM |
Number of companies listed until 31/12/2022 |
58 |
Financial companies |
(12) |
Companies that did not disclose complete data over the study period |
(05) |
Total |
41 |
Source: Based on BRVM reports.
3.2. Data Processing Method
Our study is based on quantitative data using the activity reports of listed companies. Various statistical and econometric methods and techniques are applied sequentially, as follows.
Firstly, the descriptive statistics of the sample in terms of the number of observations, mean, standard deviation, minimum and maximum.
Secondly, a correlation test is performed to check the relationship between the independent variables. If two independent variables are highly correlated, reflecting a reasonably perfect value for the correlation coefficient (around 1.0), the model used may present a multicollinearity problem, in which case one independent variable must be removed. Also, if the correlation test between the dependent and independent variables shows a zero coefficient, then there is no correlation between the variables studied. Consequently, this independent variable cannot explain the research model.
Thirdly, we will test for multi-collinearity. Multi-collinearity occurs when an independent variable is highly correlated with one or more other independent variables in a multiple regression equation. Multi-collinearity is a problem because it statically compromises the significance of an independent variable. It is therefore important that the VIF predictor is calculated. If the VIF is greater than 5, there is a multi-collinearity problem.
Fourthly, the Fisher test is used to make the best choice between the models. The choice between fixed effect and random effect will depend on the results of the tests carried out. The test results are presented in the empirical results section.
3.3. Operationalization of Variables and Econometric Models
3.3.1. Operationalization of Variables
The aim of this study is to understand the role played by the quality of accounting and financial information in the link between dividend policy and investment policy. Two independent variables (dividend policy and quality of accounting information) and one dependent variable (investment efficiency) as well as three control variables (indebtedness, economic performance and financial performance) were used. The choice of these different variables stems from the important role they play in corporate decision-making and their use in previous studies.
The indicator most commonly used to assess the quality of accounting and financial information in previous studies is Jones’ (1991) discretionary accruals model. This model is widely used in recent studies, notably Bampoky and Houewou (2023). In this study, we have adopted the same formula.
The variable relating to dividend policy (DIV) has been addressed by several authors in previous studies. These include Yenny Dwi Handayani & Ewing Yuvisa Ibrani (2023) have used different indicators such as earnings per share, payout ratio and retention rate. In our study, we looked at dividend policy in terms of the ratio between the dividend per share and the share price.
The investment policy variable (INVEFF) was captured by reference to studies by Elaoud & Jarboui (2015), André et al., (2014) and Bampoky and Houewou (2023), using the residuals from the investment model presented by these authors. The operationalization of the variables is summarized in Table 2 below.
Table 2. Definition and measurement of variables.
Variables |
Abbreviation |
Measures |
Efficiency of investment |
INVEFF |
Absolute value of the residual from the André et al., (2014) and Bampoky & Houewou (2023) models Multiplied by −1; INVEFF = −|δit| |
Quality of accounting information |
QIC |
Bampoky & Houewou model (2023) QICit = −|εit| |
Dividend policy |
DIV |
Dividend per share divided by share price |
Economic Performance |
ROA |
Ratio of gross operating profit to total assets |
Financial Performance |
ROE |
Net income/equity ratio |
Debt |
END |
The ratio of financial debt to total assets |
Interaction |
QIC*DIV |
Moderating variable |
3.3.2. Business Models
Taking into account the methodology adopted, the following models are used to test the moderating effect of the quality of accounting information on the relationship between dividend policy and the effectiveness of the investment policy of companies listed on the BRVM.
Model 1: Relationship between dividend policy and investment efficiency
INVEFFit = β0 + β1DIVit + β2QICit + β3ENDit + β4ROAit + β5ROEit + εit (1)
Model 2: Interaction between the quality of accounting information and the relationship between dividend policy and investment efficiency
INVEFFit = β0 + β1DIVit + β2QICit + β3ENDit + β4ROAit + β5ROEit
+ β6DIVit * QICit + εit (2)
4. Results and Discussion
The presentation of the results of our study in this section covers descriptive statistics, statistical tests and multiple regressions.
4.1. Descriptive Statistics of the Study
The descriptive statistics for the companies in our sample are shown in the table above. These statistics relate to certain parameters such as the average, standard deviation, minimum and maximum.
Table 3 shows that the companies in our sample recorded an average decline of 4.5% in terms of investment efficiency. This explains the improvement in investment-related decision-making among companies listed on the BRVM. With reference to the studies by Elaoud & Jarboui (2015) and Bampoky and Houewou (2023), who calculated investment efficiency as the absolute value of residuals multiplied by −1, a higher value means higher efficiency (INVEFF = −|δit|). The companies in our sample pay out an average of 15% of their profits as dividends to shareholders. Some companies did not pay any dividend, which brings the minimum value of this variable to zero (0) compared with a maximum value of 81%. The variable quality of accounting information records an average value of −2.5%, which shows a low value and consequently a low quality of accounting information, as stated in previous studies by Bampoky & Houewou (2023), the error term (εit) is an inverse measure of the quality of accounting information. εit is low, the higher the quality of the accounting information (−|εit| = QICit).
Table 3. Descriptive statistics for the study.
Variables |
Obs. |
Mean |
Std.Dev |
Min |
Max |
INVEFF |
615 |
−0.045 |
0.164 |
−2.930 |
−0.108 |
DIV |
615 |
0.150 |
0.144 |
0.000 |
0.810 |
QIC |
615 |
−0.025 |
0.035 |
−0.051 |
−0.016 |
END |
615 |
0.129 |
0.075 |
0.015 |
0.350 |
ROA |
615 |
0.047 |
0.843 |
−0.900 |
0.780 |
ROE |
615 |
0.155 |
0.260 |
−0.765 |
0.801 |
Source: Based on our estimates.
The average level of debt (END) is 12.9%, with a maximum value of 35%, which is still less than 50%. The mobilization of external resources by companies is a major problem. Average economic performance is 4.7%, compared with 15.5% for financial performance.
4.2. Correlation Matrix of Independent Variables
The correlation matrix is important in quantitative studies to highlight existing correlations between the variables in the study. This matrix provides information on the conditions required to carry out multi-collinearity tests and to see which model is suitable for the study in terms of significance. The table below shows the results obtained.
The results in Table 4 show that some of the independent variables are correlated with each other and no variable has a zero correlation; consequently, no variable should be removed from the model. This correlation between variables may lead to a multi-collinearity problem, which we will check using the Variance Inflation Factor (VIF) test. The table below shows the results of the multi-collinearity test.
Table 4. Correlation matrix between independent variables.
|
DIV |
QIC |
END |
ROA |
ROE |
DIV |
1.000 |
|
|
|
|
QIC |
0.065 |
1.000 |
|
|
|
END |
0.034 |
0.035 |
1.000 |
|
|
ROA |
0.299 |
−0.029 |
0.062 |
1.000 |
|
ROE |
0.038 |
0.053 |
0.036 |
0.084 |
1.000 |
Source: Based on our estimates.
4.3. Multi-Collinearity Test and Model Specification
Multi-collinearity is a problem that can arise with multiple regression analysis. It refers to a situation where two or more independent variables are highly correlated with each other, resulting in a paradoxical effect. This problem is detected using the Variance Inflation Factor (VIF) test. When multi-collinearity is present, one of the highly correlated variables must be removed from the regression model. The results of the multi-colinearity test are shown in Table 5 below.
Table 5. Results of multi-collinearity of variables.
Variables |
VIF |
1/VIF |
DIV |
1.11 |
0.900 |
QIC |
1.21 |
0.826 |
END |
1.09 |
0.917 |
ROA |
1.25 |
0.80 |
ROE |
1.35 |
0.740 |
Mean VIF |
1.202 |
|
Source: Based on our estimates.
From the analysis of Table 5, the average value of the variance inflation factor (VIF) of our model is 1.202 and remains below the norm (2) and clearly shows that there is no multi-collinearity at stake in our research model (Hoang et chu 2013).
The fact that the model is free of multi-collinearity problems suggests that the results are consistent with the advantage of using panel data, which is characterized by less collinearity.
The autocorrelation test is also performed using the Woodbridge test on panel data. The test result shows an F (1.449) = 2.149; Prob > F = 0.1434. This result implies that there is no autocorrelation in the model. After checking the problems of correlation, autocorrelation and multi-collinearity, we tested the relevance of our regression model. The results showed that the Fixed effect model is more appropriate. The regression results are presented in Table 6 below.
Table 6. Regression results for the two models.
|
Indicators |
Model 1 |
Model 2 |
DIV |
Coefficient |
−0.0052 |
−0.0030 |
t-value |
−4.84 |
−2.72 |
Sig.level |
0.000 |
0.053 |
QIC |
Coefficient |
0.0015 |
0.018 |
t-value |
−5.86 |
1.93 |
Sig.level |
0.001 |
0.026 |
END |
Coefficient |
−0.0001 |
−0.0323 |
t-value |
0.080 |
1.16 |
Sig.level |
0.938 |
0.000 |
ROA |
Coefficient |
0.0189 |
0.059 |
t-value |
1.93 |
1.70 |
Sig.level |
0.053 |
0.089 |
ROE |
Coefficient |
0.057 |
0.019 |
t-value |
−3.499 |
2.01 |
Sig.level |
0.000 |
0.000 |
DIV*QIC |
Coefficient |
- |
0.020 |
t-value |
- |
1.54 |
Sig.level |
- |
0.000 |
Con. |
Coefficient |
0.2456 |
0.4552 |
t-value |
7.97 |
6.71 |
Sig.level |
0.000 |
0.000 |
Model |
F-statistic |
133.12 |
150.37 |
p-value (F-statistic) |
0.000 |
0.000 |
R-squared |
0.4564 |
0.7517 |
Source: Based on our estimates.
According to Table 6, Model 1 captures the influence of each independent variable on the dependent variable through the efficiency of the investment.
INVEFFit = 0.2456 − 0.0052DIVit + 0.0015QICit − 0.0001ENDit
+ 0.0189ROAit + 0.057ROEit + εit (1)
The estimation results for model 1 show that the dividend (DIV) has a negative impact on investment efficiency (at a significance level of 1%). This result indicates that a dividend distribution reduces investment opportunities for listed companies. When companies pay out 1% of their earnings in the form of a dividend, investment efficiency is reduced by 0.52%. This result is in line with studies conducted by Nguyen (2022) and Brav et al. (2005) and Uzomah & Ihe (2021), which found a negative relationship between dividend policy and investment efficiency in developed and dynamic markets. For these authors, companies listed on the Vietnam stock market retain more liquidity for investment and pay less dividend. They confirm that when a company decides to pay a higher percentage of dividends, it eliminates access to future investment opportunities because it has less retained earnings available to finance itself.
The quality of accounting information (QIC) is an essential element in a company’s decision-making process. It has a positive and significant influence on the investment policy of companies listed on the BRVM. Good quality accounting information improves investment efficiency by 0.15%. This result confirms the work of Houewou (2022) and Harymawan et al. (2020), who found that the quality of accounting information has a positive impact on the efficiency of investment policy.
The decision to invest cannot be made without mobilizing financial resources. An investment is an expenditure made by a company with the aim of acquiring means of production in order to generate future cash flows. The mobilization of this external financial resource provides enormous advantages for the company, with reference to the financial leverage effect in modern economies. In our study, the debt (END) shows a negative and significant influence (1%) on the investment policy. Thus, for a 1% mobilization of external resources, the efficiency of investment is reduced to −0.01%. This result corroborates the work of Houewou (2022), who states that the managers of listed companies are somewhat reluctant to resort to external financing because of the fear of losing their position as managers. The reluctance of managers to take on debt may increase the risk of the company going bankrupt, forcing them to reduce their debts while cutting back on optimal investments. This situation potentially increases the prospect of under-investment, which has a negative impact on the effectiveness of investment decisions.
Economic performance (ROA) and financial performance (ROE) have a positive influence on the effectiveness of investment decisions by companies listed on the BRVM. For every 1% improvement in economic and financial performance, investment efficiency improves by 1.89% and 5.7% respectively. This shows that good profitability enhances the value of companies through efficient investment.
Our results are in line with the work of Houewou (2022), Nguyen (2022), who find that overinvestment is negatively associated with firm performance and that performance positively influences investment efficiency. Also Suhadak et al., (2019) in their study also result in a positive and significant relationship between profitability and firm value. The latter states that a good profitability improves the value of the company through a good use of financial resources intended for profitable investment.
For the appropriation of the model, the regression results show a value of R2 = 0.4565, which shows that the model’s independent variables could explain 45.65% of the variation in investment efficiency (INVEFF).
Model 2 as shown in the equation above:
INVEFFit = 0.4552 − 0.081DIVit + 0.018QICit − 0.023ENDit
+ 0.059ROAit + 0.019ROEit + 0.020DIVit * QICit + εit (2)
Understands the role of the quality of accounting information in the relationship between dividend policy and the investment policy of companies listed on the BRVM. The estimation of this model shows that the moderator variable (DIV*QIC) is positively correlated at the 5% significance level with the investment efficiency of companies listed on the BRVM. This result confirms the studies by Ramalingegowda et al., (2013); Houewou (2022) and Nguyen (2023). The relationship between dividend policy and investment efficiency improved significantly under the moderating effect of the quality of accounting information. We note −0.0052 for (model 1) and −0.0030 for (model 2). This shows that the quality of accounting information is important in the relationship between dividend policy and investment efficiency. The results confirm the agency theory, the bird-in-the-hand theory, the signals theory and the asymmetry of information. The positive influence of accounting information quality as a moderating variable in the relationship between dividend policy and investment efficiency shows that the effect of dividends on investment is less negative for companies with higher accounting information quality than for those with lower accounting information quality.
For the appropriation of the model, the regression results show a value of R2 = 0.7517 This shows that the independent variables of the model could explain 75.17% of the variation in investment efficiency (INVEFF). Furthermore, the R2 of model 2 (R2 = 0.7517) is higher than the (R2 = 0.4565) of model 1 so we can deduce that the moderating effect is established according to Baron and Kenny (1986).
5. Conclusion
Previous studies carried out on European stock markets show that dividend policy has a negative influence on companies’ investment decisions. The quality of accounting information improves this negative influence between dividend policy and investment decision. Our study on African stock markets, in particular the Bourse Regional des Values Mobilières, complements this work by analyzing the moderating role of the quality of accounting information in the relationship between dividend policy and investment efficiency.
The results of our estimations show that the quality of accounting information positively moderates the relationship between dividend policy and investment efficiency of companies listed on the BRVM. Indeed, the quality of the accounting information of listed companies provides a degree of confidence to investors, who prefer to receive a small dividend while allowing companies to invest more in order to increase the return on equity and the value of the company. This confirms the positive effect of the dividend policy on investment efficiency, which is conditioned by the moderating role of the quality of accounting information.
Like all scientific work, our study is not without its limitations. Compared with other previous studies, which look at dividend policy in terms of three variables, in our case we looked at it in terms of a single variable measured by the ratio between the dividend per share and the share price on the stock market. Other variables such as the payout ratio and the retention rate can also be analyzed.